Is it possible to mirror an AE interface that has child interfaces on different VC members to a single port on the following VCs:
Of course I understand that the bandwidth of the AE interface could not exceed the bandwith of the analyser port for all traffic to be mirrored.
I would like to reset some QFX3500's that my company just got in back to fresh out of box factory config, however they seem to be stuck in a Virtual-chassis setup. I have even tried doing a fresh clean install from the loader prompt with the --format, and these units still seem to be stuck in the virtual chassis config.
When I try doing the command "show chassis hardware detail" I get the following output.
root@P6893-C:LC:0% cli
root@P6893-C> show chassis hardware detail
error: the chassis-control subsystem is not running
root@P6893-C>
I have tried the "restart chassis-control" command. The units just come back stating that they could not be started.
I have tried doing a zeroize and going into config and doing a "load factory-default" when I do this command I get the following output.
root@P6893-C# load factory-default
warning: activating factory configuration
[edit]
'protocols'
warning: statement has no contents; ignored
[edit]
root@P6893-C#
for the life of me I just can't get these units to reset to factory defaults.
What am I missing here?
Hi everybody,
I hope you are doing fine, have a few questions:
1) Let say we have Virtual chasis with Sw1,SW2,SW3 with following roles:
SW1 is master
SW2 is back up RE
Sw3 is line card
Can Sw1 and SW2 forward tproduction traffic though they have assumed Master and back up RE role? Most likely they can as it is very similar to VDX fabric technology in Brocade, just want to be sure as I have never configured EX switch in Virtual chasis .
Thanks and have a nice day!!
Hi all,
We have couple EX4300 switches in VC configuration with ELS software version 14.1X53-D35.3.
They are used as a distribution switches and I'm trying to simplify current DHCP relay configuration.
I'm coming from Cisco world where you would only need to define ip helper-address under vlan interface to make DHCP relay configuration work.
With our current confguration everything is working but Junos is creating this subscriber database or binding database, whatever is right term.
And it also creates these Access-internal routes to routing table for each individual DHCP binding.
I would like to get rid of this behavior and make setup as simple as possible, like in Cisco.
forward-snooped-clients configured-interfaces;
overrides {
allow-snooped-clients;
}
server-group {
windowsservers-group {
10.10.2.1;
10.10.2.3;
}
}
group office-pc {
active-server-group windowsservers-group;
interface irb.20;
}
group office-voip {
active-server-group windowsservers-group;
interface irb.21;
}
When installing EX switches we discovered that we needed these options so that lease renewing would work:
forward-snooped-clients configured-interfaces;
overrides {
allow-snooped-clients;
And this had to be configured same way to our core switches where our DHCP servers are hosted. Core switches are also EX4300 with just different port configuration.
So, question is: What configuratio would stop Junos creating DHCP binding database and work just like Cisco does? So that unicast DHCP renew requests would go thru to server also.
Hello,
I have virtual chassis with two members:
Member ID Status Serial No Model prio Role Mode Mode ID Interface 0 (FPC 0) Prsnt xxxx ex4200-48t 129 Master* N VC 1 vcp-0 1 vcp-1 1 (FPC 1) Prsnt xxxx ex4200-24t 129 Backup N VC 0 vcp-0 0 vcp-1
I can not understand how to clean up /var partition, already did all from KB, but still have alarm.
Any suggestion?
admin@hostname> show chassis alarms
1 alarms currently active
Alarm time Class Description
2010-01-10 22:22:07 EET Minor Host 1 /var partition usage is high
{master:0}
admin@hostname> request session member 1
--- JUNOS 15.1R5.5 built 2016-11-25 16:39:56 UTC
warning: This chassis is operating in a non-master role as part of a virtual-chassis (VC) system.
warning: Use of interactive commands should be limited to debugging and VC Port operations.
warning: Full CLI access is provided by the Virtual Chassis Master (VC-M) chassis.
warning: The VC-M can be identified through the show virtual-chassis status command executed at this console.
warning: Please logout and log into the VC-M to use CLI.
{backup:1}
admin@hostname> start shell user root
Password:
root@hostname:BK:1% df -h | grep var
/dev/da0s3e 123M 101M 12M 90% /var <<<<<<<<
/dev/da0s3d 369M 17M 323M 5% /var/tmp
/dev/md10 118M 19M 89M 18% /var/rundb
root@sw-01:BK:1% pwd
/var
root@hostname:BK:1% du -chs * | sort -rn
240K mfs
210K etc
138K run
112K jail
102K db
100K etcroot
41M total <<<<<<<<
19M rundb
17M tmp
14K spool
12K home
10K jails
6.0K at
4.2M log
4.0K transfer
4.0K sw
4.0K root
4.0K cron
4.0K crash
4.0K BSD.var.dist
2.0K yp
2.0K validate
2.0K rwho
2.0K preserve
2.0K named
2.0K msgs
2.0K mail
2.0K logical-systems
2.0K heimdal
2.0K empty
2.0K chroot
2.0K bin
2.0K backups
2.0K accountThanks.
Hi everybody
Let say we have three EX 4200 switches: SW1, SW2,SW3, we want to use " Dynamic Installation method" to configure them in a chasis mode.
We cable VCP ports in daaisy chained ring toplogy. We set up SW1 and SW2 priority as 255, while SW3 has a default priority of 128.
We power SW1 first then SW2 and finally SW3. They assume following roles:
SW1--Elected master
SW2--Back up RE
SW3--Line card
Let say we power down SW1, Sw2 assume master and SW3 BACK UP RE role now, next we power on SW1. What role SW1 ,SW2 and SW3 assume ?
MASTER ELECTION CRITERIA:
The algorithm proceeds from the top condition downward until the stated condition is satisfied:
This is my hunch based on the above:
When SW1 powers up and join the Chasis, SW3 will assume LINE CARD ROLE because it has the lowest priority of 128.
SW1 and SW2 have the same priority of 255 so 2nd criteria is used i.e who has the more time in the chasis, since SW1 just joins Chasis therefore SW2 remains Master and SW1 assumes BACK UP RE role.
Am I correct?
Thanks and have a nice weekend!!
Hi everybody,
1)Let say we have two EX4200 switches as master RE and back up RE in a virtual chasis mode, we make some configuration changes on master and simly commit it ( we do not use commit synchronize option), will the change still be propagated to back up RE? Is there any minimum interval or it happens right away as we make configuration changes to Master RE?
2) Is it required all CONFIGURATION CHANGES be entered via MASTER RE i.e we can not go to a member switch and try to make configuration changes there ?
Thanks and have a nice weekend!!
Hi everybody.
Let say we have two switches SW1,SW2 which do not have dediacted VCP port, we therefore decided to use UPLINK port as VPC port.
Both switches have factory default config. Before we can connect them via VCP ports we must configure uplink port as VCP port.
We start with SW1:
SW1:
SW1> request virtual-chassis vc-port set pic-slot 1
Above I did not specify any member id as we do not have one( remember this SW has factory defaults). So what value is assumed above?
SW2:
SW2>request virtual-chassis vc-port set pic-slot 1
Above I did not specify any member id as we do not have one( remember this SW has factory defaults). So what value is assumed above?
Thanks and have a nice weekend!!
Hi All,
As i'm undertstand QFX5100 need additional license if we want make VCF. But may i know how about normal VC. Is it QFX support it without license?
Thanks and appreciate anyone feedback
A while ago we discovered that qfx10k is dropping transit frames with triple vlan tags. After long investigation with JTAC, A-JTAC, DEV Team, etc... they confirmed the problem but they are refusing to accept it as a bug. Moreover, The DEV team has confirmed that this is expected behavior, i.e. if there is a triple tagged packet incoming to the switch(which is definitely a valid ethernet frame), the packet shall be dropped and not allow to pass through. As a result we have completely unusable switch with no chance to get it fixed.
Does anyone has experience how to deal in such situation? Any suggestions (except the court) are appreciated
Thanks in advance
Best Regards
Michael Varyan
Hi,
We are going to improve our current network by adding additional switch, router and firewall in order to eliminate single point of failure. Below is the network topology (also as attached), but I am not quite sure on the design especially on the redundant trunk group between routers and switches. Appreciate if you guys could give some comments on it. Secondly, does J-Series router support LACP? Our internet router is using Cisco 2960 and will connect to 2xJ6350 routers using LACP.
Hi all,
I see lots of output drops on a 30g AE (consisting of three 10g interfaces). It's a 2 member EX4550 VC:
Model: ex4550-32f
Junos: 15.1R5.5
AE detail stats:
Output errors:
Carrier transitions: 44, Errors: 0, Drops: 127061556, MTU errors: 0, Resource errors: 0
Queue counters: Queued packets Transmitted packets Dropped packets
0 0 8786218535053 126988388
1 0 0 0
5 0 0 0
7 0 29847977601 73168
Thing is, at max the interface is doing 'only' ~ 8Gbps. This is only at night. During the day (like now) it's only doing approx. 2G in and 2G out. And even than I got ~ 254 drops within a minute.
I understand tail-dropped means they're dropped because the buffer is full. But I really can't figure out why. At least looking at the bandwith, there's enough room.
One thing I just noticed, while checking when it appeared to have started: ~ 2016-12-29 01:36:04
That night, we replaced this stack with a new one which was installed with 15.1R5.5; (we did this to reduce downtime, as it's a core stack)
Also, from that moment on there's a huge increase of outbound traffic.. that concerns me ..
So one could think it's caused by the JunOS version?
Any insights would be appreciated! 
[edit: added images]
Hi,
I am looking into turning an EX3300 switch into a Virtual Chassis with 2 additional switches, making it 3 switches all together.
I've found many articles online explaining how to do this however I cannot find whether it requires downtime on the switch that currently exists?
Can someone advise whether when configuring the active switch into a Virtual Chassis master, does this require downtime?
I know adding additional switches to an existing VC does not require downtime, but not sure if when setting up a new VC.
Thank you,
Vishal
So in the PR for R16 it shows that juniper stops support on VC. I am needing to move from R14.X to R16.X so that I can finally have LRm optical support for my client that has a surplus of LRMs. We were looking into EVPN to migrate too but with all the case studys and burn in surrounding the new 10ks and the issues they are having we are going to hold off on going EVPN right away so we are looking at doing VRRP. I dont see any documents on this migration.
I'm installing a new set of QFX5100s and am having difficulty with the vme management interface. I have two QFX5100s in a VC pair, and have C0 (em0) on each QFX5100 connected to our management network. Very minimal config on the QFXs, and we're at least initially not doing any layer-3 on them.
admin@dcas01a-ts1dc> show interfaces terse vme
Interface Admin Link Proto Local Remote
vme up down
vme.0 up down inet 10.38.4.87/24
{master:0}
admin@dcas01a-ts1dc> show interfaces terse em0
Interface Admin Link Proto Local Remote
em0 up up
em0.0 up up
{master:0}
Hi ,
Does there have any document description for how many IRB interface supported on EX switch ?
Like:
Cisco Nexus 3K suppoed 1450 SVI
Extreme X670 suppoed 2000 SVI
Thanks for your answer .
Hello,
I have a little issue related to the MAC learning by a standalone switch. My setup is very simple - two end stations connected to one switch, MAC dynamically learned without any issues, pings are ok. Once I connected a new switch to some free ports on switch under question, no trunks yet, no STP either the MACs start bouncing in the table - learning of my PCs MAC is constantly changing . I mean I learn my macs from other UP ports those that will be used for trunking bust still access ports (default config). I dont observe that behaviour on cisco 3750.
Simple bridging within one ex4200 became impossible as soon as I turn up other ports. I'm using default config at the moment. RSTP temporarly disabled/deleted though. It seems I experience MAC flooding-relearning within the box L2 table. Ping now is intermittent. If more ports connected the hosts then will not be able to reach each other.
Please advise what is missing, I think not much as I wanted this switch to perform basic L2 forwarding between local access ports.
I use Juno 12.3R5.7
Hi!
I have expirienced with a weird behaviour our EX4600 switch. I add sfp+ module at port and see that this port becomes up without any cables. What is a reason?
All leds on port are green
sfp - 10GBASE SR MM JDSU PLRXPL-SC-S43-J1 850 nm
1> show interfaces xe-1/0/23
Physical interface: xe-1/0/23, Enabled, Physical link is Up
Interface index: 795, SNMP ifIndex: 690
Description: Rezerv-SAPHANA
Link-level type: Ethernet, MTU: 1514, MRU: 0, Speed: 10Gbps, BPDU Error: None, MAC-REWRITE Error: None, Loopback: Disabled,
Source filtering: Disabled, Flow control: Disabled, Media type: Fiber
Device flags : Present Running
Interface flags: SNMP-Traps Internal: 0x4000
Link flags : None
CoS queues : 12 supported, 12 maximum usable queues
Current address: f4:a7:39:8d:e3:1a, Hardware address: f4:a7:39:8d:e3:1a
Last flapped : 2017-07-21 17:12:37 GMT-3 (00:02:47 ago)
Input rate : 0 bps (0 pps)
Output rate : 0 bps (0 pps)
Active alarms : None
Active defects : None
Interface transmit statistics: Disabled
Logical interface xe-1/0/23.16386 (Index 591) (SNMP ifIndex 691)
Flags: Up 0x4004000 Encapsulation: ENET2
Input packets : 0
Output packets: 47
> show interfaces diagnostics optics xe-1/0/23
Physical interface: xe-1/0/23
Laser bias current : 5.688 mA
Laser output power : 0.5910 mW / -2.28 dBm
Module temperature : 48 degrees C / 118 degrees F
Module voltage : 3.3230 V
Receiver signal average optical power : 0.0055 mW / -22.60 dBm
Laser bias current high alarm : Off
Our Configuration had 3 EX 4200 48P switches in stack during a recent power failure the entire stack was rest and all the configuation was lost. We were able to restore the master with the backup configuration but the remaining 2 switches are stuck on loading junos. We can console into the switches and both of them are labeled as linecard 0 one was configured as backup and other was configured as line card with no split detection enabled. and virtual chassis preprovisioned. We tried to factory reset the switch so as to reconfigure it to be added to the stack but that does not work either.Tried to commit a config but keep getting error message
error: could not open /var/etc/pam.conf+: Operation not permitted
error: commit failed: foreign file propagation failed during preprocessing
When i ran the snapshot command there is no output
{linecard:0}
root> show system snapshot media internal
{linecard:0}
root>
Are the switches Dead? We dont have any JTAC support on these switches. We are going to try booting the device via USB to see if that helps us to fix the issue.
